1004 The Journal of Human Resources
Stewart and Swaffield 1997, for Britain. This reduces people’s leisure time—some- times without financial compensation. If not taken into account statistically, a neg-
ative relationship between the unemployment rate and life satisfaction could thus reflect either depressed salaries or reduced leisure time after economic shocks.
The above-mentioned effects on salaries and working hours refer to realized con- sequences. However, high unemployment also affects anticipated economic distress,
as, for instance, the probability that a worker may himself experience a spell of unemployment in the future increases. Other people’s unemployment might thus
primarily affect people through the information conveyed about potential hazards and not through social comparisons.
5
Moreover, people also may expect salary de- creases, reduced promotion opportunities, fewer possibilities to change occupations,
etc. In the remainder of the paper, we use the term economic insecurity when ad-
dressing the psychic costs of negative anticipatory feelings due to both worries and fear about job loss or, alternatively, an income reduction in the future and to the
many consequences that might follow like reduced social status, loss of a social network, necessary adjustments in consumption habits, etc..
III. The Life Satisfaction Gap Between Employees in the Public and the Private Sectors
To study the importance of the effects of high unemployment on individuals’ economic insecurity independent of general effects on society, we
compare the subjective well-being of workers in the public and private sectors. This approach hinges on specific assumptions and conditions with regard to the quality
of the subjective well-being data as well as the characteristics of the labor market.
A. Subjective Well-Being as a Proxy for Individual Welfare
For the question at hand, the validity of the subjective well-being method depends on three main conditions:
1 Subjective well-being scores contain information on the respondent’s global evaluation of his or her life. It is necessary in other words, that reported attitudes
are not merely arguments in the utility function, or a subutility function, as Kimball and Willis 2006—in our mind, rightly—conjecture for measures of current affect.
The problem of only analyzing a subutility function holds for all empirical measures to a greater or lesser degree. Here, data on reported satisfaction with life is used
that is understood to refer to the cognitive component, the rational or intellectual
5. Social comparisons are a prevalent issue in economic happiness research. There is, however, only limited work on social comparisons in the unemployment domain for example, Clark 2003. We are not aware
of any research on social comparisons of employed people with their unemployed “peers.” Two counter- vailing factors are likely to operate. On the one hand, there is an information, signaling or fear effect the
main argument in our paper. On the other hand, there is the classical comparison effect in terms of relative standing as mentioned in the main text. This latter effect would counteract some of the general negative
consequences of unemployment on society.
Luechinger, Meier, and Stutzer 1005
aspects of subjective well-being Lucas, Diener, and Suh 1996. Behind the score indicated by a person lies a cognitive assessment of the extent to which their overall
quality of life is judged in a favorable way Veenhoven 1993. This includes—in our context—hard-to-measure aspects such as general concerns about the state of
the economy, or anxiety about crime rates or job losses. Based on this, we assume that the standards underlying people’s life satisfaction judgments are sufficiently
close to those that the individual would like to pursue in order to maximize welfare.
2 Measurement error for reported subjective well-being is not correlated with the variables of interest. Schwarz and Strack 1999 document that well-being reports
are susceptible to the ordering of questions, the wording of questions, and actual mood, for example. In our main analysis based on the GSOEP, we use a question
of overall life satisfaction. Throughout the panel, this question is asked at the end of the questionnaire after a bloc of questions on marital status and family relations.
In this setting, we see no indication for the labor market situation priming subjective well-being responses.
3 Reported life satisfaction contains sufficient information relative to noise about actual individual welfare that statistical research is fruitful. There is substantial
evidence for this. Measures of reported subjective well-being passed a series of validation exercises: They correlate with behavior and aspects generally associated
with people’s happiness. Reliability studies have found that reported subjective well- being is moderately stable and sensitive to changing life circumstances. Consistency
tests, for instance, reveal that happy people are more likely to be rated as happy by friends and family members for references, see Frey and Stutzer 2002b; Clark,
Frijters, and Shields al. 2008.
B. Sectoral Differences in Job Security
